DESCRIPTION (Applicant's abstract): Events leading to emphysema are poorly understood in humans, but likely include structural and physiologic alteration characterized by inflammatory processes, arteriolar wall thickening, and parenchymal destruction. Early processes occur with conventional clinical tests remaining normal. Using humans and animals, we have developed objective, reproducible CT-based measures for the characterization of pulmonary parenchyma (the adaptive multiple feature method - AMFM), and for the measurement of pulmonary blood flow parameters. We will utilize a well-defined animal model of emphysema to allow for the correlation of these CT measures with lung morphometry, for correlating the CT measures during the evolution of the emphysema with lung morphometry, and for providing well-defined lung morphometric changes to train the AMFM. Through a well-designed prospective study, we will also study human subjects over a range of smoking history, age, and lung function results. Parenchymal perfusion and parenchymal tissue characteristics will be assessed using these dynamic high-resolution volumetric CT data. These CT-based measures will be applied to the whole lung and to lung lobes and will enable in vivo structural and functional properties to be measured. The overriding hypothesis is that pulmonary emphysema is characterized by an essential sequence of pathophysiologic events that can be ultimately assessed using high speed HRCT. These events are regional inflammatory events including a blood flow response followed by alveolar destruction. On completion this study will relate quantitative regional pathology to quantitative regional lung CT structure and quantitative CT measures of regional perfusion in the normal dog lung and in the dog lung exposed to elastase digestion. In the normal, the cigarette smoking, and the aging human, the study will relate these quantitative regional lung CT structure and perfusion measures to each other, to lung function tests, to an NO index of inflammation, and to the tissue patterns associated with known pathology obtained from the dog studies. Such information is critical for the further understanding of smoking-related disease processes in the human lung as well as providing for objective measures of the severity of, and distribution of, smoking-related effects. In the human population, without such measures of pathophysiology and outcomes, other mechanistic studies and tests of newly emerging interventions have not and will not be practically achievable.